Objective: This study aims to introduce and assess a novel AI-driven tool developed for the classification of orthodontic arch shapes into square, ovoid, and tapered categories. Methods: Between 2016 and 2019, we collected 450 digital dental models. Applying our inclusion and exclusion criteria, we refined our dataset to 50 models, ensuring a focused and detailed analysis. Plaster casts were digitized into 3D models with AutoScan-DS-EX. Three trained evaluators then measured mesiodistal and arch widths using MeshLab. The development of DentalArch was undertaken in two versions: the first version incorporates 18 input parameters, including mesiodistal widths (from the first molar to the first molar, totaling 14) and arch widths (1 intercanine, 2 interpremolar, and 1 intermolar, totaling 4); the second version uses only 4 parameters related to arch widths. Both versions aim to predict the arch shape. An evaluation of 28 machine learning methods through a k = 5-fold cross-validation was conducted to determine the most effective techniques. Results: In the tests, the performance evaluation of the DentalArch software in detecting arch shapes revealed that version 1, which analyzes 18 parameters, achieved an accuracy of 94.7% for the lower arch and 93% for the upper arch. The more streamlined version 2, which assesses only four parameters, also showed high precision with an accuracy of 93.0% for the lower arch and 92.7% for the upper arch. Conclusions: DentalArch provides a tool with potential use in orthodontic diagnostics, particularly in the task of arch shape classification. The software offers a less subjective and data-driven approach to arch shape determination. Moreover, the open-source nature of DentalArch ensures its global availability and encourages contributions from the orthodontic community.

Context: Accurate dental arch shape prediction is crucial for orthodontic treatment and personalized dental appliance creation. This study introduces a computer vision-based tool for predicting arch shapes in 3D dental models. Objective: To automate the selection of dental arch shapes through mathematical model analysis. Method: A dataset of 484 digital dental models was narrowed to 50 through specific criteria. Experts classified these into ovoid, square, and tapered shapes using 3M templates. An automated 3D dental arch shape prediction tool was developed, incorporating automatic alignment, cusp detection, curve fitting with a sixth-order polynomial, and model comparison. Our validations employed attribute agreement analysis, the root mean squared error, the sum of squared errors, and a Gage R&R Study. Results: This study achieved a 90% agreement rate in the evaluator vs. standard comparison for the lower jaw, as well as 78% for that of the upper jaw. The Gage R&R study confirmed measurement reliability, and the sixth-order polynomial model was identified as optimal for arch shape description. The tool’s predictive accuracy was validated through comparative analysis. Conclusion: This research introduces an effective automated method for selecting dental arch shapes. The tool demonstrated substantial accuracy, with the potential to significantly enhance orthodontic diagnostic and treatment planning processes. Future research could further refine this methodology by exploring advanced mathematical models and incorporating machine learning techniques to optimize the selection process.

La enseñanza tradicional de la programación presenta dificultades en el desarrollo de pensamiento algorítmico y lógico. Para enfrentar esto, se ha explorado el aprendizaje activo, que involucra directamente al estudiante en la resolución de problemas. El Tecnológico de Antioquia implementó un experimento con una herramienta de aprendizaje activo para enseñar fundamentos de programación a estudiantes de estratos socioeconómicos diversos, buscando superar las dificultades en la comprensión de conceptos como algoritmos y estructuras de control. Evaluar experimentalmente el impacto de una herramienta de aprendizaje activo, basada en ejercicios medidos por siete dimensiones, en comparación con métodos tradicionales en la enseñanza de la programación. Se realizó un estudio con estudiantes de primer semestre del Tecnológico de Antioquia. La muestra se dividió en un grupo de control, que recibió enseñanza tradicional, y un grupo experimental, que utilizó una herramienta de aprendizaje activo con más de 1000 ejercicios en Python. Se aplicaron tres pruebas diagnósticas (inicial, intermedia y final) para medir el progreso en siete dimensiones del pensamiento computacional. Los datos se analizaron con Power BI. Resultados: Se observó una mejora en el rendimiento del grupo experimental entre la primera y segunda prueba diagnóstica, especialmente en Pensamiento Lógico y Reconocimiento de Patrones. El análisis del etiquetado manual de 1018 ejercicios de la herramienta arrojó que el 59.86% fueron correctamente generados de forma automática. Las dimensiones con menor rendimiento inicial fueron Resolución de Problemas, Algoritmos y Pensamiento Lógico. La implementación de modelos activos de programación, apoyados por herramientas interactivas, muestra ser efectiva para mejorar el aprendizaje en entornos universitarios. La herramienta demostró ser un enfoque inclusivo, efectivo en estudiantes de diversos perfiles socioeconómicos, aunque se requiere refinar la automatización en la generación de ejercicios para optimizar su calidad y reducir la revisión manual.